[sdiy] Communications with voices in a polyphonic synth

karl dalen dalenkarl at yahoo.se
Thu Mar 11 01:31:27 CET 2010


As Niel already have suggested, UART in 9 bit mode and a simple
protocol handler who deals with the 7bit mode for parameter are
a robust solution, latency will be low.

Spam the Uart lines with parameter commandos on all voice cards
will most likely bottleneck the lines at some point, but instead
of assuming massive parameter transmissions analyze how many 
parameters will actually be sent by the user? You will be suppriced
that a typical Disco music set (Bony M whatever) will be quite few
filter frequency, perhaps a stomping LFO driving whatever. It also
depends on the voice architecture and what synthesis methods you
have implemented and what kind of music the user actually does.
In other words dont overdo your design for things that will most
likely never be used or seldom used.

I2C has some High speed (Mhz) modes dont know if the dSpic uses
these but it seams cumbersome with the handshaking back and fort.

If you need very high speed you could use I2s if the manufacturer
has implement the multi device and interuppt driven part in the
transceiver section.I recall SGS did on their Cortex MCU?!

Just took a peek in the 128gp manual and it has I2S multdevice
high speed implemented, it also has a buffer on transmit and
receive so this will most likely do for your app.Look for the
DCI section in the manual.

Another are a TDM interfaces whos speciffically designed for
this kind of tasks but you will most likely only find it in
plain DSP chips.

Reg
KD    

__________________________________________________
Använder du Yahoo!?
Är du trött på spam?  Yahoo! E-post har det bästa spamskyddet som finns 
http://se.mail.yahoo.com 




More information about the Synth-diy mailing list